NVIDIA GB10 Grace Blackwell vs NVIDIA RTX A6000
Side-by-side comparison of specifications, performance, and efficiency.
NVIDIA GB10 Grace Blackwell
Architecture
Blackwell
Year
2025
VRAM
128 GB
TDP
180 W
NVIDIA RTX A6000
Architecture
Ampere
Year
2020
VRAM
48 GB
TDP
300 W
Specifications
NVIDIA GB10 Grace Blackwell NVIDIA RTX A6000
Architecture
Blackwell Ampere
Launch Year
2025 2020
Form Factor
Integrated PCIe
VRAM
128 GB48 GB
Memory Bandwidth
—768 GB/s
TDP
180 W300 W
Process Node
— 8nm
| Spec | NVIDIA GB10 Grace Blackwell | NVIDIA RTX A6000 |
|---|---|---|
| Architecture | Blackwell | Ampere |
| Launch Year | 2025 | 2020 |
| Form Factor | Integrated | PCIe |
| VRAM | 128 GB | 48 GB |
| Memory Bandwidth | — | 768 GB/s |
| TDP | 180 W | 300 W |
| Process Node | — | 8nm |
Performance (TFLOPS)
NVIDIA GB10 Grace Blackwell NVIDIA RTX A6000
FP64
—0.6 TFlops
FP32
31.0 TFlops38.7 TFlops
FP16
—38.7 TFlops
BF16
—38.7 TFlops
FP8
1000.0 TFlops—
INT8
—77.4 TFlops
| Precision | NVIDIA GB10 Grace Blackwell | NVIDIA RTX A6000 |
|---|---|---|
| FP64 | — | 0.6 TFlops |
| FP32 | 31.0 TFlops | 38.7 TFlops |
| FP16 | — | 38.7 TFlops |
| BF16 | — | 38.7 TFlops |
| FP8 | 1000.0 TFlops | — |
| INT8 | — | 77.4 TFlops |
FLOPS by Precision
Efficiency — TDP vs FP32 Performance
Analysis
Most Efficient
NVIDIA
GB10 Grace Blackwell
0.17 TFLOPS/W
Get Comparison Updates
New GPUs added weekly. Be the first to see how they compare.